ABSTRACT Osteoarthritis (OA), a debilitating age-related disease associated with pain, stiffness and poor functioning is a major risk factor for mobility disability. Although early osteoarthritic changes within the joint commence during mid-life (40-65 years of age), early detection of disease is limited given the lack of robust and reliable OA biomarkers. Currently detection relies upon costly imaging modalities. Late detection of OA compromises the opportunity for early intervention and prevention of disease progression, leaving only symptom management or, ultimately, joint replacement as strategies for treatment. Recent evidence and scientific appreciation that underlying metabolic dysfunction is a risk factor for osteoarthritis incidence and progression suggests that biomarkers which identify individuals with disordered metabolism may be relevant for subclinical markers of OA. Metabolomics, a newly evolving field, analyzes small molecules (metabolites) in biological specimens. Metabolomics analysis has successfully identified novel biomarkers for diagnosis, monitoring and treatment for age-related diseases such as prostate cancer, diabetes and stenosis and autoimmune diseases such as rheumatoid arthritis. A small but growing number of studies in animal and human populations have reported that metabolomics yields potential biomarkers with good discrimination between OA patients and normal controls including metabolites associated with collagen, branched chain amino acid, energy, and tryptophan metabolism. However, no studies to date have neither used metabolomics to identify biomarkers for OA incidence nor evaluated biomarkers among individuals matched for age and body size. We propose to conduct a metabolomics analysis of osteoarthritis risk within the longitudinal Michigan Study of Women's Health Across the Nation (MI-SWAN). Specifically, 63 MI-SWAN women who developed radiographic knee OA during follow-up will be age- and BMI-matched with 63 MI-SWAN women who remained OA-free during follow- up. Banked plasma specimens from baseline (when all subjects were OA-free) will be used to conduct metabolomics analyses using the targeted lipids eicosanoids platform (Aim 1) which includes profiles from 28 eicosanoids, the lipidomics platform (Aim 2) which profiles lipids from over 10 classes including 431 unique lipid species, and an untargeted platform (Aim 3) which profiles at least 250 known compounds to identify candidate biomarkers for knee osteoarthritis risk. Relative quantitation of these metabolites will be compared within the matched pairs of women who did and did not develop incident knee OA during follow-up. This K01 award will provide needed training and skill development in metabolomics, the associated bioinformatics considerations, and translation to clinical care and yield preliminary data to support the submission of an R01 application. This training will enable the candidate to develop as an independent investigator providing leadership in the application of metabolomics research in aging with the long-term goal of applying this approach to identify key metabolic pathways involved in aging and the disablement process.